Application of a neural network model for OCT image classification in the diagnosis of choroidal melanoma
https://doi.org/10.21516/2072-0076-2026-19-2-100-108
Abstract
Early and accurate diagnosis of choroidal melanoma (CM) is required to initiate treatment in a timely manner and improve the prognosis. The purpose of the study was to develop an automated system for supporting medical decision-making in the diagnosis of choroidal tumors based on optical coherence tomography (OCT) images using artificial intelligence (AI) technologies and to evaluate its performance. Material and methods. The article describes the creation process and evaluates the effectiveness of the YOLO11n-cls convolutional neural network in the automatic classification of OCT fundus images into three classes: CM, choroidal nevus, and healthy fundus. Results. To train and evaluate the model, a dataset of 1,700 OCT images obtained at the Helmholtz National Medical Research Center of Eye Diseases over the period 2014–2024 was created. Training was performed using augmentation methods and weights pre-trained on ImageNet. According to the testing results, the accuracy of the model was 95 %. The primary limitations identified include insufficient variability in unique cases and the necessity for further external validation of the model. Conclusion. The potential of using AI-based technologies in the creation of programs to support medical decision-making in ophthalmological practice has been demonstrated.
About the Authors
E. B. Myakoshina
Helmholtz National Medical Research Center of Eye Diseases
Russian Federation
Russian Federation
Elena B. Myakoshina — Dr. of Med. Sci., senior researcher of ocular oncology and radiology department, assistant professor, chair of eye diseases.
14/19, Sadovaya-Chernogryazskaya St., Moscow, 105062
S. V. Saakyan
Helmholtz National Medical Research Center of Eye Diseases; The Russian University of Medicine
Russian Federation
Russian Federation
Svetlana V. Saakyan — Corresponding member of the Russian Academy of Sciences, Dr. of Med. Sci., professor, head of ocular oncology and radiology department, Helmholtz National Medical Research Center of Eye Diseases; head of the academic department of chair of eye diseases of N.A. Semashko Nanional Research Institute of clinical medicine, The Russian University of Medicine.
14/19, Sadovaya-Chernogryazskaya St., Moscow, 105062; 4, Dolgorukovskaya St., Moscow, 127006
A. O. Ukina
Gatchina Interdistrict Clinical Hospital
Russian Federation
Russian Federation
Anastasia O. Ukina — ophthalmologist, hospital ophthalmology department.
15A, building 1, Roshchinskaya St., Gatchina, Leningrad region, 188300
D. D. Garri
Limited Liability Company “Artificial Networks and Technologies”
Russian Federation
Russian Federation
Denis D. Garri — PhD student, chair of eye diseases of the faculty of continuing professional education, specialist in working with medical data.
30, Kozlov St., Moscow, 121357
References
1. Myakoshina E.B. Small choroidal melanoma and pseudomelanomas: methods of differential diagnostics (literature review). Part 3. Russian ophthalmological journal. 2020; 13 (4): 91–8 (In Russ.). https://doi.org/10,21516/2072-0076-2020-13-4-91-98
2. Miakoshina E.B., Saakian S.V. Optical coherence tomography in diagnostics of small choroidal melanoma. Vestnik oftal’mologii. 2020; 136 (1): 56–64 (In Russ.). https://doi.org/10,17116/oftalma202013601156
3. Brovkina A.F., Stoyukhina A.S., Musatkina I.V. Diagnostic potential of optical coherence tomography for small choroidal melanomas. Vestnik oftal’mologii. 2016; 132 (5): 23–34 (In Russ.). https://doi.org/10,17116/oftalma2016132523-34
4. Stoyukhina A.S., Musatkina I.V. Morphogenesis of choroidal melanomas in OCT imaging. Vestnik oftal’mologii. 2018; 134 (5–2): 186–94 (In Russ.). https://doi.org/10,17116/oftalma2018134051186
5. Saakyan S.V., Neroev V.V., Yurovskaya N.N., et al. Optical coherence tomography of tumor-associated retinal changes in neoplasm of choroid. Russian ophthalmological journal. 2009; 2: 35–41 (In Russ.).
6. Neroev V.V., Saakyan S.V., Myakoshina E.B., et al. Classification of tumor-associated retinal changes in uveal tumors. Russian ophthalmological journal. 2010; 4: 35–9 (In Russ.).
7. Garri D.D., Saakyan S.V., Khoroshilova-Maslova I.P., et al. Мethods of machine learning in ophthalmology: Review. Ophthalmology in Russia. 2020; 17 (1): 20–31 (In Russ.). https://doi.org/10,18008/1816-5095-2020-1-20-31
8. Kermany DS, Goldbaum M, Cai W, et al. Identifying medical diagnoses and treatable diseases by image-based deep learning. Сell. 2018 Feb 22; 172 (5): 1122–31.e9. doi: 10,1016/j.cell.2018.02.010
9. Tang F, Wang X, Ran AR, et al. A Multitask deep-learning system to classify diabetic macular edema for different optical coherence tomography devices: A multicenter analysis. Diabetes Care. 2021 Sep; 44( 9): 2078–2088. doi: 10.2337/dc20-3064
10. Selvam A, Ong J, Bollepalli SC, et al. Artificial intelligence in choroid through optical coherence tomography: A comprehensive review. Authorea Preprints. 2023. doi: 10,36227/techrxiv.24076470,v2
11. Katalevskaya E.A., Sizov A.Yu., Gilemzianova L.I. Artificial intelligence algorithm for segmentation of pathological structures on optical coherence tomography scans. Russian journal of telemedicine and electronic health care. 2022; 8 (3): 21–7 (In Russ.). doi: 10,29188/2712-9217-2022-8-3-21-27
12. Neroev V.V., Bragin A.A., Zaytseva O.V. Diagnostics of retinal pathologies by optical coherence tomography images using artificial intelligence tools. Russian ophthalmological journal. 2023; 16 (3): 47–53 (In Russ.). https://doi.org/10,21516/2072-0076-2023-16-3-47-53
13. Patel RH, Foltz EA, Witkowski A, Ludzik J. Analysis of artificial intelligence-based approaches applied to non-invasive imaging for early detection of melanoma: a systematic review. Cancers (Basel). 2023 Sep 23; 15 (19): 4694. doi: 10,3390/cancers15194694
14. Tailor PD, Kopinski PK, D’Souza HS, et al. Predicting choroidal nevus transformation to melanoma using machine learning. Ophthalmol Sci. 2024 Jul 20; 5 (1): 100584. doi: 10,1016/j.xops.2024.100584
15. Karamanli K-E, Maliagkani E, Petrou P, Papageorgiou E, Georgalas I. Artificial intelligence in decoding ocular enigmas: A literature review of choroidal nevus and choroidal melanoma assessment. Applied Sciences. 2025; 15 (7): 3565. https://doi.org/10,3390/app15073565
16. Ragab MG, Abdulkadir SJ, Muneer A, et al. A comprehensive systematic review of YOLO for medical object detection (2018 to 2023). IEEE Access. 2024; 12: 57815–536. doi: 10,1109/ACCESS.2024.3386826
17. Katalevskaya E.A., Katalevskiy D.Yu., Tyurikov M.I., Velieva I.A., Bolshunov A.V. Future of artificial intelligence for the diagnosis and treatment of retinal diseases. Russian journal of clinical ophthalmology. 2022; 22 (1): 36–43 (In Russ.). doi: 10,32364/2311-7729-2022-22-1-36-43
18. Shulga LA, Saakyan SV, Skladnev DA. Finding a uniform system of retina coefficient for early detection of initial stage of pathological changes. In: Ng EYK, Acharya UR, Tamura T, eds. Distributed Diagnosis and Healthcare. Vol. 3. American Scientific Publishers; 2011: 321–5.
19. Aldughayfiq B, Ashfaq F, Jhanjhi NZ, Humayun M. YOLO-Based deep learning model for pressure ulcer detection and classification. Healthcare (Basel). 2023 Apr 25; 11 (9): 1222. doi: 10,3390/healthcare11091222
20. Pinto-Coelho L. How artificial intelligence is shaping medical imaging technology: A survey of innovations and applications. Bioengineering (Basel). 2023 Dec 18; 10 (12): 1435. doi: 10,3390/bioengineering10121435
21. Ali ML, Zhang Z. The YOLO framework: A comprehensive review of evolution, applications, and benchmarks in object detection. Computers. 2024; 13 (12): 336. https://doi.org/10,3390/computers13120336
Review
For citations:
Myakoshina E.B., Saakyan S.V., Ukina A.O., Garri D.D. Application of a neural network model for OCT image classification in the diagnosis of choroidal melanoma. Russian Ophthalmological Journal. 2026;19(2):100-108. (In Russ.) https://doi.org/10.21516/2072-0076-2026-19-2-100-108
Views:
102
JATS XML
Email this article 